Skip to content


Using scream-detecting drones to save disaster victims

In disaster situations, rescuers often only have a matter of minutes to find and save potential victims.

 In this race against time, drones are the perfect tool to support emergency response efforts as they can cover large areas of unstable or dangerous terrain without putting rescue workers at risk. Now, researchers at Germany's Fraunhofer FKIE have developed new technology that allows drones to listen out for screams to rapidly pinpoint exactly where victims are located.We caught up with Macarena Varela, one of the researchers behind the project, to discuss her recent presentation of this breakthrough at the 180th Meeting of the Acoustical Society of America and how she hopes her drones will help save lives.



Morressier: Your research has received a great deal of media interest since you recently presented it at the 180th Meeting of the Acoustical Society of America (ASA). How does it feel to see your work resonating so much with the public? 

Macarena Varela: It's been crazy since my presentation. The thing is, I’ve been talking about this for years and dreaming that it would be possible - that we could really fly with a drone and microphones and be able to find the positions of people in need. I really believe in this project, but was not expecting this much positive resonance. I'm really happy that I got the attention of the media as this may also attract people to work with us and it may allow me to focus more of my time on this research.

Morressier: Was the ASA Annual Meeting the first place that you presented your findings?

Varela: Yes, they got the premiere, including a demonstration of the drone flying and carrying the microphones and everything.

Morressier: What was the experience of virtually presenting your poster at the ASA meeting like? 

Varela: The main difference is that I had to record my presentation. It was limited to 12 minutes, so it had to be very compact. When I’m presenting in front of an audience, I can see the time and if I hesitate, it's not such a big deal. But it’s a bit different when it's recorded. I think I recorded the presentation five or six times as it was my first time recording myself and I had to get used to it. I recorded using Zoom, as Morressier suggested. But after it was recorded, it was actually quite nice as I didn't have to worry much about the presentation anymore, I just had to focus on answering any questions. The whole experience was positive and I think next time I’ll have some practice and be able to record my presentation quicker. 

Morressier: How does presenting and attending a conference virtually compare to in-person? What are the pros and cons?

Varela: I do find that nicer to be there in person. Because I get to know people, to talk with them and get to know more about their work. Someone presents and then you always have these gatherings afterwards. You can go to the person directly and ask questions in an informal setting, so you get to know people in a different context. And then sometimes as a result of these gatherings you establish new collaborations. That element I think is missing in a virtual setting.

Morressier: So to go back to your research, can you walk me through the process to get where you are today? 

Varela: We initially developed a bigger system that was able to locate different kinds of sounds. We were able to localize specific sounds and enable this system to be placed on a car or a vehicle to be used in disaster recovery. These microphones can hear much further than people and can detect the impulsive sounds that people make when they’re trapped or injured, such as screams, so they can locate the rescue position and transmit this to the emergency group. 

A few years ago one of my co-authors, Dr. Wirth, started to study MEMS microphones, the type that you find in cell phones. They’re tiny, however he noticed that their quality and robustness sensitivity was comparable to conventional condenser microphones. That’s when we came up with the idea of using these microphones in a crowsnest array and placing them on a drone to support disaster recovery. Before, only analogue condenser microphones were used, and they are far too heavy to be carried on a drone. We are now transferring all the knowledge that we have from this bigger system to the smaller one. We are still in the testing phase, but we’ve already been able to filter out the noise of the drone in order to extract and locate the interesting signals very precisely.

Morressier: What sparked your initial interest in researching how to use drones in disaster recovery?

Varela: The major advantage of drones is that you can cover a very large area in a very short period of time, even dangerous or hard to reach areas. In such a situation, people could be heavily injured and need to be rescued as fast as possible. Every minute counts in saving their lives. That's why using drones to locate people makes so much sense.

Morressier: After presenting your initial research at the ASA Annual Meeting is your aim to publish your findings in a journal as a next step?

Varela: Yes, I definitely plan on publishing our findings so that other researchers can add their ideas and build on our work or maybe even cooperate with us.

Astronaut on blue background, the benefits of technology partnerships